An efficient solution method for integrated unit commitment and natural gas network operational scheduling under “Duck Curve”

Pouria Maghouli | Farhad Fallahi

---

URL :   http://research.shahed.ac.ir/WSR/WebPages/Report/PaperView.aspx?PaperID=137906
Date :  2020/10/21
Publish in :    International Transactions on Electrical Energy Systems
DOI :  https://doi.org/https://doi.org/10.1002/2050-7038.12662
Link :  http://dx.doi.org/https://doi.org/10.1002/2050-7038.12662
Keywords :Duck Curve, Renewable energy source, integrated power and gas system, Gas network, power to gas, Benders decomposition

Abstract :

---

In this paper, we study an integrated unit commitment and natural gas network operational scheduling with renewable energy (wind and photovoltaic power plants) which changes the load curve to a duck curve. Wind and photovoltaic energy are both uncertain in nature. The availability of photovoltaic energy at certain times of day changes the pattern of the load. This load curve which is called “Duck Curve”, raises the question of whether the power system could provide the necessary ramp rate and capacity for full utilization of photovoltaic energy. The use of gas power plants to deal with the sharp ramp rate would have a negative impact on the flow of gas in the gas network. That is, the uncertain output of renewable resources requires the flexibility of gas power plants which affects the gas network. Storage devices of electricity and gas, line-pack and the technology of electricity to gas transformation (P2G) could be effective solutions for reducing fluctuations and ensuring reliable operation of electricity and gas networks. The underlying problem is known to be a complex non-convex problem. Compared to existing modeling and solution methods, we propose some efficient reformulation which enables us to rewrite the problem as a convex mixed integer program using Benders decomposition. Given the convex reformulation, we can solve the resulting convex model to optimality. We demonstrate the effectiveness of the proposed method by solving some standard examples such as the modified IEEE 24-Bus test system and the Belgian 20-node gas network.